Please explain in detail!! (feel free to include hand written work as well as st
ID: 3069202 • Letter: P
Question
Please explain in detail!! (feel free to include hand written work as well as stata commands)
1. The tables below displays data from a case-control study on smoking and invasive cervical cancer with data stratified by number of sexual partners 0-1 Partners Smoker Non-smoker Total Cervical Cancer 12 25 37 No Cervical Cancer 21 118 139 No Cervical Cancer 142 150 292 Total 143 176 2 or More Partners Smoker Non-smoker Total Cervical Cancer 96 92 188 Total 238 242 480 . cc Cancer Smoke Ere count], by (NumPart) NumPart OR [95% conf . Interval] M-H weight 0-1 2.697143 1.102266 1.058383 6.612562 2.982955 (exact) 2 or More 7512169 .61748 27.21667 (exact) 1.080447 2.130222 Crude M-H Combined 1.517697 1.259799 exact) 900359 1.762734 Test of homogeneity (M-H) chi2 (1) - 3.74 Pr>chi2 0.0531 Test that combined OR1: Mantel-Haenszel chi2(1) 1.86 0.1728 a. Is there enough evidence to suggest that the relationship between smoking and cervical cancer is different for people with 0-1 partners than it is for people with 2 or more partners at the 0.05 significance level? Explain your answer thoroughly Does number of sexual partners confound the relationship between smoking status and invasive cervical cancer? Explain your answer thoroughly b. Based on your answers to a. and b., report the most appropriate measure(s) of association from the Stata output. c. d. Overall, is there a statistically significant relationship between smoking and invasive cervical cancer at the 0.05 significance level? Justify your answerExplanation / Answer
a) The Mantel Haensel test of Homogeneity of odds ratio is not statistically signifcant (chi2 with df 1=3.74, p-value=0.0531) at the 0.05 significance level. The hypothesis is fail to rejected. Conclude that there is no statistically significant evidence that the association of smoking with the event of cervial cancer is not different depending on the no. of partners status.
b) In order to see the confounding, we need to calculate the relative difference
Relative Difference= (|1.259799-1.517697|)/1.517697*100= 16.99%
The % relative difference is nearly 17%, sugesting no confounding.
c) The Mantel Haensel test of Homogeneity of odds ratio is not statistically signifcant (chi2 with df 1=3.74, p-value=0.0531) at the 0.05 significance level. The hypothesis is fail to rejected. Conclude that there is no statistically significant evidence that the association of smoking with the event of cervial cancer is not different depending on the no. of partners status.
d) The result shown here indicate that, on average, there is no association between smoking and invasive cervical cancer(chi2 on df=1 = 1.86, p-value=0.1728).